In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
In [2]:
path='E:/AA DS TM/2-Time Series Data Analysis/individual_stocks_5yr'
company_list=['AAPL_data.csv','GOOG_data.csv','MSFT_data.csv','AMZN_data.csv']
all_data=pd.DataFrame()
for file in company_list:
curr_df=pd.read_csv(path+'/'+file)
all_data=pd.concat([all_data,curr_df])
all_data.shape
Out[2]:
In [3]:
all_data.head()
Out[3]:
Analysing closing price of all the stocks¶
In [4]:
tech_list=all_data['Name'].unique()
In [5]:
all_data.dtypes
Out[5]:
In [6]:
all_data['date']=pd.to_datetime(all_data['date'])
In [7]:
all_data.dtypes
Out[7]:
In [8]:
plt.figure(figsize=(20,12))
for i,company in enumerate(tech_list,1):
plt.subplot(2,2,i)
df=all_data[all_data['Name']==company]
plt.plot(df['date'],df['close'])
plt.xticks(rotation='vertical')
plt.title(company)
Analysing the Total Volume of the stock being traded each day¶
In [9]:
import plotly.express as px
In [10]:
for company in tech_list:
df=all_data[all_data['Name']==company]
fig=px.line(df,x='date',y='volume',title=company)
fig.show()
Analysing Daily price change in stock¶
In [11]:
df=pd.read_csv('E:/AA DS TM/2-Time Series Data Analysis/individual_stocks_5yr/AAPL_data.csv')
df.head()
Out[11]:
In [12]:
df['Daily_Price_change']=df['close']-df['open']
df.head()
Out[12]:
In [13]:
df['1Day % return']=((df['close']-df['open'])/df['close'])*100
In [14]:
df.head()
Out[14]:
In [15]:
fig=px.line(df,x='date',y='1Day % return',title=company)
fig.show()
Analysing the monthly mean of the close feature¶
In [16]:
df2=df.copy()
In [17]:
df2.dtypes
Out[17]:
In [18]:
df2['date']=pd.to_datetime(df['date'])
In [19]:
df2.set_index('date',inplace=True)
In [20]:
df2.head()
Out[20]:
In [21]:
df2['2013-02-08':'2013-02-13']
Out[21]:
In [22]:
df2['close'].resample('M').mean().plot()
Out[22]:
In [23]:
df2['close'].resample('Y').mean().plot(kind='bar')
Out[23]:
In [ ]:
Analysing whether the stock prices of the Tech co's are correlated or not¶
In [24]:
aapl=pd.read_csv('E:/AA DS TM/2-Time Series Data Analysis/individual_stocks_5yr/AAPL_data.csv')
aapl.head()
Out[24]:
In [25]:
amzn=pd.read_csv('E:/AA DS TM/2-Time Series Data Analysis/individual_stocks_5yr/AMZN_data.csv')
amzn.head()
Out[25]:
In [26]:
msft=pd.read_csv('E:/AA DS TM/2-Time Series Data Analysis/individual_stocks_5yr/MSFT_data.csv')
msft.head()
Out[26]:
In [27]:
goog=pd.read_csv('E:/AA DS TM/2-Time Series Data Analysis/individual_stocks_5yr/GOOG_data.csv')
goog.head()
Out[27]:
In [28]:
close=pd.DataFrame() # creating an empty dataframe to store all the close values
In [29]:
close['aapl']=aapl['close']
close['amzn']=amzn['close']
close['msft']=msft['close']
close['goog']=goog['close']
In [30]:
close.head()
Out[30]:
In [31]:
import seaborn as sns
In [32]:
sns.pairplot(data=close)
Out[32]:
In [33]:
sns.heatmap(close.corr(),annot=True)
Out[33]:
Analysing the daily return of each stock & their correlation¶
In [34]:
aapl.head()
Out[34]:
In [35]:
data=pd.DataFrame()
In [36]:
data['aapl_change']=((aapl['close']-aapl['open'])/aapl['close'])*100
data['amzn_change']=((amzn['close']-amzn['open'])/amzn['close'])*100
data['msft_change']=((msft['close']-msft['open'])/msft['close'])*100
data['goog_change']=((goog['close']-goog['open'])/goog['close'])*100
In [37]:
data.head()
Out[37]:
In [38]:
sns.pairplot(data=data)
Out[38]:
In [39]:
sns.heatmap(data.corr(),annot=True)
Out[39]:
Value at Risk analysis for tech co's¶
In [40]:
sns.distplot(data['aapl_change'])
Out[40]:
In [41]:
data['aapl_change'].std()
## 68% of entire data lies btw -1.8 to 1.8
Out[41]:
In [42]:
data['aapl_change'].std()*2
## 95% of data lies btw -2.37 to 2.37
Out[42]:
In [43]:
data['aapl_change'].std()*3
## 99.7% data lies btw -3.56 to 3.56
Out[43]:
In [44]:
data['aapl_change'].quantile(0.1)
Out[44]:
In [45]:
data.describe().T
Out[45]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]: